Abstract: Based on the DNDC（Denitrification-Decomposition Model）model and incorporated continuous monitoring data， such as CO₂ flux， from the experimental area， the study investigated the effects of different nitrogen forms on CO₂ emissions from saline-alkali grasslands in the Hexi Corridor， and evaluated the sensitivity of CO₂ emissions to climate change， soil properties， and management practices. The results showed that， compared to no nitrogen addition （CK）， the application of different nitrogen sources （organic nitrogen， ON； inorganic nitrogen， IN； mixed fertilization， MN） significantly increased soil CO₂ emissions by 24.27% to 71.03% during the growing season. Furthermore， model evaluation results indicated that the NRMSE （Normalized Root Mean Square Error） values ranged from 0.117 to 0.221， and the R² values ranged from 0.605 to 0.906， suggesting that the DNDC model could effectively simulate the impact of different nitrogen inputs on CO₂ emissions. Sensitivity analysis revealed that the annual average temperature was the most critical factor influencing CO₂ emissions. Based on climate scenario projections for the next 50 years （2024—2073）， compared to CK， different nitrogen inputs were expected to increase CO₂ emissions by 18.73% to 54.37%. This study provides important references for the local improvement and applicability verification of the DNDC model， offering theoretical support for reducing greenhouse gas emissions from saline-alkali grasslands in the Hexi Corridor under the context of global climate change.

Key words: CO₂ emissions, DNDC model, Different nitrogen input forms, Sensitivity analysis

摘要： 本研究以河西盐渍化草地为研究对象，基于反硝化-分解模型 （Denitrification-Decomposition Model，DNDC），结合试验区连续监测数据CO₂排放通量等数据，探讨了不同形态氮输入对盐渍化草地CO₂排放的影响，并评估了气候变化、土壤特性和管理措施对CO₂排放的敏感性。研究结果表明，相较于不添加氮肥（CK），不同氮源（有机氮ON、无机氮IN、混合施肥MN）添加显著增加了土壤CO₂排放量，且生长季CO₂排放量增加了24.27%~71.03%。此外，模型评估结果显示，归一化均方根误差值在0.117~0.221之间，R²值介于0.605~0.906之间，表明DNDC模型能够有效模拟不同氮源输入对CO₂排放的影响。敏感性分析表明，年平均温度是影响CO₂排放的最关键因素。而基于未来50年（2024—2073年）的气候情景模型预测表明，相较于CK，不同形态氮输入将使得CO₂排放量增加18.73%~54.37%。该研究为DNDC模型的局部改良及适用性验证提供了重要参考，并为在全球气候变化背景下，河西走廊盐渍化草地减少温室气体排放提供理论支持。

关键词: CO₂排放, DNDC模型, 不同形态氮输入, 敏感性分析